Two candidates score 68% on a readiness test. One reads "68", feels sick, and buys another question bank. The other spends ninety minutes on the breakdown, finds that endocrine and health promotion account for most of the loss and that a third of the misses were rushed rather than unknown, and rewrites the next three weeks accordingly.
Same number, completely different value extracted. The readiness test is a diagnostic instrument and diagnostics only pay off if you read the whole report.
When to take one
Early enough to act on it. A predictor exam taken four days before your board date can only frighten you — there is no time left to change anything. Useful timing:
- Week 1 — a cold baseline, before studying. Uncomfortable, and the single most informative block you will take.
- Around week 7 — mid-plan, to check whether the build phase moved the domains you targeted.
- Week 9 or 10 — a full-length simulation under real conditions, with time to respond to what it shows.
Three is plenty. Taking a predictor every week burns the material, and you start remembering items rather than reasoning through them, which inflates the score and destroys the instrument.
Conditions matter or the data is worthless
A readiness test taken on the sofa, paused twice, with a textbook nearby, measures nothing you can use. If you want the number to mean something:
- Full length, one sitting, timed.
- Phone in another room.
- No lookups, no pausing, no rewinding.
- Same time of day as your booked exam.
- Somewhere that is not your most comfortable chair.
Reading the report: the four-cause analysis
The score tells you how much you lost. The item review tells you why, and there are only four reasons. Go through every missed question and tag it:
| Cause | What happened | The fix |
|---|---|---|
| Content gap | You did not know the material | Study that topic. This is the only cause where more content is the answer. |
| Retrieval failure | You knew it but could not produce it under time | Active recall and spaced repetition, not rereading |
| Misread stem | You answered a question that was not asked | Read the last line first; identify the verb before the options |
| Rushed or fatigued | A careless error you would not make fresh | Pacing practice and stamina blocks |
Now count each column. That distribution decides the next three weeks. Mostly content gaps means back to the build phase. Mostly retrieval failures means your study method is too passive. Mostly misread stems means a mechanical fix worth several points for almost no effort. Mostly fatigue errors means longer blocks, not more content.
Read the rationales for correct answers too
Some of what you got right, you guessed. Those items are gaps that the score does not show, and they are the ones most likely to reappear as misses on the real exam. If the rationale surprises you, log it.
Turning the breakdown into a plan
Take the domain-level percentages and put them next to the blueprint weights from your certifying body's test content outline. Two columns, one simple product:
Priority = (100 − your domain score) × blueprint weight
Sort descending. The top of that list gets your time, regardless of how you feel about the topics on it. A 55% in a domain worth 25% of the exam outranks a 30% in a domain worth 4% — comfortably — even though the 30% is the number that stings.
What the number does and does not tell you
Readiness tests correlate with real outcomes well enough to be useful. They are not prophecies, for three concrete reasons:
- Different scale. A percentage on a practice bank is not the scaled score the boards report, and the two do not map onto each other cleanly.
- Different item mix. Commercial banks vary in difficulty and in how closely they follow the current blueprint.
- Different you. The version of you who took it in week 7 is not the version sitting the exam in week 12 — assuming you actually change the plan.
Use it as a compass. A low score early is information, not a sentence. A high score late is encouraging, not a guarantee, and it is a poor reason to stop drilling.
If the score is bad
Give yourself a day, then do the four-cause analysis. Almost always it reveals that the problem is narrower than it felt — two or three domains plus a mechanical issue, rather than everything. Rebuild the plan around exactly those, and take the next predictor no sooner than three weeks later so there is something to measure.
If the score is good
Two failure modes here. The first is stopping: a strong result five weeks out is not a finish line, and skills decay. The second is complacency about the domains that were merely adequate — the ones sitting just above your target line are where a bad day takes points, and they deserve the drilling that the disaster domains get automatically.
Key points
- Take a readiness test early enough that the result can still change your plan.
- The domain breakdown is the deliverable; the overall percentage is almost decoration.
- Analyse why you missed each item, not just which ones — four different causes need four different fixes.
- Predictor scores correlate with outcomes but do not determine them; treat them as a compass, not a prophecy.